CNC milled Waveguide for the D2608/913000 (Peerless HDS)

Hello!

I am planning a two-way speaker with the ScanSpeak D2608/913000 tweeter in a waveguide and a ScanSpeak 22w/4534g00. Therefore, I made a test-piece on my CNC-machine, and am asking for comments on the frequency responses at different angles.

The sidewalls of the waveguide enlarge on the contour of a circle with 225 mm diameter. The waveguide itself has a diameter of 225 mm and is 36 mm deep. Find attached an image of the waveguide construction and the prototype.

The measurements shown are gated, unsmoothed, measured at 1 m distance, and at angles from 0 to 90 degrees from top to bottom.

I am a bit concerned about the lack of directionality between 2 and 8 kHz. The 0, 15 and 30 degree measurements are a bit bundled up. What do you think?

Thanks for any comments!
 

Attachments

  • construction.PNG
    construction.PNG
    164.4 KB · Views: 300
  • A6408292.jpg
    A6408292.jpg
    575.2 KB · Views: 276
  • A6408293.jpg
    A6408293.jpg
    360.8 KB · Views: 275
  • A6408294.jpg
    A6408294.jpg
    277.5 KB · Views: 300
  • meas.png
    meas.png
    318.2 KB · Views: 283
It would be my assumption that the bunching from 2-3.5k is edge diffraction from the narrow baffle.

That could be the case. The baffle is 250 mm wide, so that makes sense. I will add chamfering at a later stage.

The bunching at 7k is a puzzler, though

I am assuming that it could be related to the 32 mm throat diameter. I have a 1 mm lip there, which I cannot really remove. There might also be a small gap between tweeter and waveguide.